Introduction

Astragalus holmgreniorum, Holmgren’s Milk-vetch, is an endangered endemic perennial herb of the Northern Mojave Desert near St. George, UT. Once the leguminous pods detach from the mother plant they begin to dry and dehice. At this stage the morphological characteristcs exhibit aerodynamic qualities. This study was performed to determine if wind dispersal is possible and to what distance pods potentially move.

Attached fleshy pods. Photo credit Susan Meyer

Attached fleshy pods. Photo credit Susan Meyer

Detached dehiscent pods. Photo credit Aly DeNittis

Detached dehiscent pods. Photo credit Aly DeNittis

Procedure

A runway was designed with different substrates: sand, small gravel, and medium gravel. Three degrees of slope were tested: no slope, upslope and downslope. Three speeds of wind (4.3 m/s, 7.1 m/s, 10.7 m/s) were provided by a fan to mimic habitat and environmental conditions in a controlled laboratory setting. For each treatment combination a new group of pods was tested totaling 324 individuals each trialed at two orientations to the wind. Each pod was timed from the start of the fan to the first stop of movement. Pods were given an equal cut off time if no movement occured.

Data Analysis

Of the 324 trials 89% of pods moved a distance and stopped within 15 seconds of wind exposure and 57% travled to 152.4 cm or more. Runway maximum length is 304.8 cm, roughly 3 meters.
The effect of wind on pod movement was evident during trials. Knowing that over 50% of our selection from the population moved a good distance in a short time is confirmation of wind as a dispersal component.

Inital investigation

Pod mass and orientation to the wind had no significant effect and the NULL hypothesis is accepted for these variables.

Analysis of variance

The table below shows significant p-values for speed, time, slope and substrate. When looking at variables combined there is significant interaction of speed, slope and substrate together.

##                               Df  Sum Sq Mean Sq F value   Pr(>F)    
## speed                          2 2027257 1013628 487.512  < 2e-16 ***
## time_s                         1 1262384 1262384 607.152  < 2e-16 ***
## slope                          4   25948    6487   3.120 0.015620 *  
## substrate                      2  221619  110810  53.295  < 2e-16 ***
## speed:time_s                   2  165555   82778  39.813 7.04e-16 ***
## speed:slope                    8   23892    2986   1.436 0.181217    
## time_s:slope                   4   17707    4427   2.129 0.077489 .  
## speed:substrate                4   58724   14681   7.061 2.02e-05 ***
## time_s:substrate               2   30766   15383   7.399 0.000744 ***
## slope:substrate                2   23743   11871   5.710 0.003727 ** 
## speed:time_s:slope             8   37326    4666   2.244 0.024675 *  
## speed:time_s:substrate         4   14266    3566   1.715 0.146818    
## speed:slope:substrate          4   44455   11114   5.345 0.000371 ***
## time_s:slope:substrate         2    4903    2451   1.179 0.309178    
## speed:time_s:slope:substrate   4    4619    1155   0.555 0.695275    
## Residuals                    270  561381    2079                     
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Here we see that substrate grade and speed interact with the pods in a significant way. On sand and small gravel the majority of pods moved a fair distance at all three speeds. On medium gravel at low and medium speed the majority of pods did not move over 100 cm and of those many did not leave the starting point.

Subset of sand, distance ~ speed

##              Df Sum Sq Mean Sq F value Pr(>F)    
## speed         2 600424  300212   167.6 <2e-16 ***
## Residuals   105 188081    1791                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Subset of small gravel, distance ~ speed

##              Df Sum Sq Mean Sq F value Pr(>F)    
## speed         2 818588  409294   95.05 <2e-16 ***
## Residuals   105 452142    4306                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Subset of medium gravel, distance ~ speed

##              Df Sum Sq Mean Sq F value Pr(>F)    
## speed         2 748941  374471   56.27 <2e-16 ***
## Residuals   105 698770    6655                   
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Conclusion/Acknowledgements

This investigation aims to fill knowledge gaps in dispersal processes of Astragalus holmgreniorum. Findings thus far have shown that substrate, speed of wind, and slope effect movement of these pods. In the right conditions the pods will move a fair distance. For the ones that moved very little or not at all from the starting point; further investigation into why is required. Continued investigation of dimensional measurements will be analyzed to see if there is any correlation of the pod curvature on distance and time.

Thanks to Dr. Susan Meyer from the US Forest Service for making this experiment possible.